Files
bespoke_stratos_0.3k/trainer_state.json

862 lines
20 KiB
JSON
Raw Permalink Normal View History

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 12.30379746835443,
"eval_steps": 500,
"global_step": 117,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.10126582278481013,
"grad_norm": 6.43825208149568,
"learning_rate": 8.333333333333333e-07,
"loss": 0.9281,
"step": 1
},
{
"epoch": 0.20253164556962025,
"grad_norm": 6.559153069315341,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.9108,
"step": 2
},
{
"epoch": 0.3037974683544304,
"grad_norm": 7.045196702684219,
"learning_rate": 2.5e-06,
"loss": 0.8874,
"step": 3
},
{
"epoch": 0.4050632911392405,
"grad_norm": 6.450857710098491,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.9102,
"step": 4
},
{
"epoch": 0.5063291139240507,
"grad_norm": 5.274518665853932,
"learning_rate": 4.166666666666667e-06,
"loss": 0.8248,
"step": 5
},
{
"epoch": 0.6075949367088608,
"grad_norm": 3.2648170762097477,
"learning_rate": 5e-06,
"loss": 0.8383,
"step": 6
},
{
"epoch": 0.7088607594936709,
"grad_norm": 2.5880252500398844,
"learning_rate": 5.833333333333334e-06,
"loss": 0.7354,
"step": 7
},
{
"epoch": 0.810126582278481,
"grad_norm": 3.845109483810076,
"learning_rate": 6.666666666666667e-06,
"loss": 0.7928,
"step": 8
},
{
"epoch": 0.9113924050632911,
"grad_norm": 4.109400755644465,
"learning_rate": 7.500000000000001e-06,
"loss": 0.7744,
"step": 9
},
{
"epoch": 1.0506329113924051,
"grad_norm": 4.311702547596846,
"learning_rate": 8.333333333333334e-06,
"loss": 0.7524,
"step": 10
},
{
"epoch": 1.1518987341772151,
"grad_norm": 3.822432777947372,
"learning_rate": 9.166666666666666e-06,
"loss": 0.6639,
"step": 11
},
{
"epoch": 1.2531645569620253,
"grad_norm": 3.5468998915875813,
"learning_rate": 1e-05,
"loss": 0.6936,
"step": 12
},
{
"epoch": 1.3544303797468356,
"grad_norm": 2.800787770940674,
"learning_rate": 9.997762161417517e-06,
"loss": 0.6682,
"step": 13
},
{
"epoch": 1.4556962025316456,
"grad_norm": 2.1669190884101175,
"learning_rate": 9.991050648838676e-06,
"loss": 0.5577,
"step": 14
},
{
"epoch": 1.5569620253164556,
"grad_norm": 2.087992110618876,
"learning_rate": 9.979871469976197e-06,
"loss": 0.5829,
"step": 15
},
{
"epoch": 1.6582278481012658,
"grad_norm": 2.7059991544282798,
"learning_rate": 9.964234631709188e-06,
"loss": 0.6041,
"step": 16
},
{
"epoch": 1.759493670886076,
"grad_norm": 2.2940983342523125,
"learning_rate": 9.944154131125643e-06,
"loss": 0.6291,
"step": 17
},
{
"epoch": 1.8607594936708862,
"grad_norm": 1.6808408200948324,
"learning_rate": 9.91964794299315e-06,
"loss": 0.6081,
"step": 18
},
{
"epoch": 1.9620253164556962,
"grad_norm": 1.6600609215129083,
"learning_rate": 9.890738003669029e-06,
"loss": 0.5927,
"step": 19
},
{
"epoch": 2.1012658227848102,
"grad_norm": 1.472819671631874,
"learning_rate": 9.857450191464337e-06,
"loss": 0.5682,
"step": 20
},
{
"epoch": 2.2025316455696204,
"grad_norm": 1.6799800601492856,
"learning_rate": 9.819814303479268e-06,
"loss": 0.5403,
"step": 21
},
{
"epoch": 2.3037974683544302,
"grad_norm": 1.3804440688611062,
"learning_rate": 9.777864028930705e-06,
"loss": 0.4855,
"step": 22
},
{
"epoch": 2.4050632911392404,
"grad_norm": 1.1432325373879264,
"learning_rate": 9.731636918995821e-06,
"loss": 0.5031,
"step": 23
},
{
"epoch": 2.5063291139240507,
"grad_norm": 0.9092916002520707,
"learning_rate": 9.681174353198687e-06,
"loss": 0.4571,
"step": 24
},
{
"epoch": 2.607594936708861,
"grad_norm": 1.1253148386623655,
"learning_rate": 9.626521502369984e-06,
"loss": 0.4672,
"step": 25
},
{
"epoch": 2.708860759493671,
"grad_norm": 1.4032465429020564,
"learning_rate": 9.567727288213005e-06,
"loss": 0.4384,
"step": 26
},
{
"epoch": 2.810126582278481,
"grad_norm": 1.1988699935589895,
"learning_rate": 9.504844339512096e-06,
"loss": 0.3985,
"step": 27
},
{
"epoch": 2.911392405063291,
"grad_norm": 0.8951456999785163,
"learning_rate": 9.437928945022772e-06,
"loss": 0.4792,
"step": 28
},
{
"epoch": 3.050632911392405,
"grad_norm": 0.9624502825360387,
"learning_rate": 9.36704100308565e-06,
"loss": 0.362,
"step": 29
},
{
"epoch": 3.151898734177215,
"grad_norm": 1.0438288012802868,
"learning_rate": 9.292243968009332e-06,
"loss": 0.3723,
"step": 30
},
{
"epoch": 3.2531645569620253,
"grad_norm": 0.9703176540776492,
"learning_rate": 9.213604793270196e-06,
"loss": 0.3369,
"step": 31
},
{
"epoch": 3.3544303797468356,
"grad_norm": 0.893163662422422,
"learning_rate": 9.131193871579975e-06,
"loss": 0.3604,
"step": 32
},
{
"epoch": 3.4556962025316453,
"grad_norm": 1.2422079407990139,
"learning_rate": 9.045084971874738e-06,
"loss": 0.3631,
"step": 33
},
{
"epoch": 3.5569620253164556,
"grad_norm": 0.9151822546049537,
"learning_rate": 8.955355173281709e-06,
"loss": 0.3553,
"step": 34
},
{
"epoch": 3.6582278481012658,
"grad_norm": 1.0114454402313395,
"learning_rate": 8.862084796122998e-06,
"loss": 0.4085,
"step": 35
},
{
"epoch": 3.759493670886076,
"grad_norm": 0.9443337928269486,
"learning_rate": 8.765357330018056e-06,
"loss": 0.2889,
"step": 36
},
{
"epoch": 3.8607594936708862,
"grad_norm": 0.865644384923322,
"learning_rate": 8.665259359149132e-06,
"loss": 0.2859,
"step": 37
},
{
"epoch": 3.962025316455696,
"grad_norm": 0.8599590553282734,
"learning_rate": 8.561880484756726e-06,
"loss": 0.36,
"step": 38
},
{
"epoch": 4.10126582278481,
"grad_norm": 0.8238766304670959,
"learning_rate": 8.455313244934324e-06,
"loss": 0.284,
"step": 39
},
{
"epoch": 4.2025316455696204,
"grad_norm": 0.8904225393629464,
"learning_rate": 8.345653031794292e-06,
"loss": 0.2657,
"step": 40
},
{
"epoch": 4.30379746835443,
"grad_norm": 0.8281971083836545,
"learning_rate": 8.232998006078998e-06,
"loss": 0.2237,
"step": 41
},
{
"epoch": 4.405063291139241,
"grad_norm": 0.9318352078173276,
"learning_rate": 8.117449009293668e-06,
"loss": 0.2345,
"step": 42
},
{
"epoch": 4.506329113924051,
"grad_norm": 0.9593192555567228,
"learning_rate": 7.99910947343957e-06,
"loss": 0.2498,
"step": 43
},
{
"epoch": 4.6075949367088604,
"grad_norm": 0.8527439601562876,
"learning_rate": 7.87808532842837e-06,
"loss": 0.2405,
"step": 44
},
{
"epoch": 4.708860759493671,
"grad_norm": 0.9473177017919994,
"learning_rate": 7.754484907260513e-06,
"loss": 0.2913,
"step": 45
},
{
"epoch": 4.810126582278481,
"grad_norm": 0.8102313751416961,
"learning_rate": 7.628418849052523e-06,
"loss": 0.286,
"step": 46
},
{
"epoch": 4.911392405063291,
"grad_norm": 0.820114624158469,
"learning_rate": 7.500000000000001e-06,
"loss": 0.2364,
"step": 47
},
{
"epoch": 5.050632911392405,
"grad_norm": 0.7927604968113404,
"learning_rate": 7.369343312364994e-06,
"loss": 0.2076,
"step": 48
},
{
"epoch": 5.151898734177215,
"grad_norm": 0.6375951563290391,
"learning_rate": 7.236565741578163e-06,
"loss": 0.1667,
"step": 49
},
{
"epoch": 5.253164556962025,
"grad_norm": 0.6651567314411381,
"learning_rate": 7.101786141547829e-06,
"loss": 0.1684,
"step": 50
},
{
"epoch": 5.3544303797468356,
"grad_norm": 0.7043980605084836,
"learning_rate": 6.965125158269619e-06,
"loss": 0.2039,
"step": 51
},
{
"epoch": 5.455696202531645,
"grad_norm": 0.7406146243351966,
"learning_rate": 6.8267051218319766e-06,
"loss": 0.1794,
"step": 52
},
{
"epoch": 5.556962025316456,
"grad_norm": 0.7560497079368976,
"learning_rate": 6.686649936914151e-06,
"loss": 0.1529,
"step": 53
},
{
"epoch": 5.658227848101266,
"grad_norm": 0.6135830262872792,
"learning_rate": 6.545084971874738e-06,
"loss": 0.1956,
"step": 54
},
{
"epoch": 5.759493670886076,
"grad_norm": 0.7112397939983846,
"learning_rate": 6.402136946530014e-06,
"loss": 0.1958,
"step": 55
},
{
"epoch": 5.860759493670886,
"grad_norm": 0.7940516397429221,
"learning_rate": 6.257933818722544e-06,
"loss": 0.217,
"step": 56
},
{
"epoch": 5.962025316455696,
"grad_norm": 0.7871752636384775,
"learning_rate": 6.112604669781572e-06,
"loss": 0.1946,
"step": 57
},
{
"epoch": 6.10126582278481,
"grad_norm": 0.5470125055488606,
"learning_rate": 5.9662795889777666e-06,
"loss": 0.1701,
"step": 58
},
{
"epoch": 6.2025316455696204,
"grad_norm": 0.5813020536626888,
"learning_rate": 5.819089557075689e-06,
"loss": 0.1426,
"step": 59
},
{
"epoch": 6.30379746835443,
"grad_norm": 0.5710246684545575,
"learning_rate": 5.671166329088278e-06,
"loss": 0.1184,
"step": 60
},
{
"epoch": 6.405063291139241,
"grad_norm": 0.8565249872894666,
"learning_rate": 5.522642316338268e-06,
"loss": 0.1978,
"step": 61
},
{
"epoch": 6.506329113924051,
"grad_norm": 0.7450325352987529,
"learning_rate": 5.373650467932122e-06,
"loss": 0.1521,
"step": 62
},
{
"epoch": 6.6075949367088604,
"grad_norm": 0.7228163461638277,
"learning_rate": 5.224324151752575e-06,
"loss": 0.1556,
"step": 63
},
{
"epoch": 6.708860759493671,
"grad_norm": 0.5931907548279679,
"learning_rate": 5.074797035076319e-06,
"loss": 0.1134,
"step": 64
},
{
"epoch": 6.810126582278481,
"grad_norm": 0.6442106835491783,
"learning_rate": 4.9252029649236835e-06,
"loss": 0.1085,
"step": 65
},
{
"epoch": 6.911392405063291,
"grad_norm": 0.6660065713778855,
"learning_rate": 4.775675848247427e-06,
"loss": 0.1296,
"step": 66
},
{
"epoch": 7.050632911392405,
"grad_norm": 0.513068792938708,
"learning_rate": 4.626349532067879e-06,
"loss": 0.1223,
"step": 67
},
{
"epoch": 7.151898734177215,
"grad_norm": 0.4707168636389631,
"learning_rate": 4.477357683661734e-06,
"loss": 0.0976,
"step": 68
},
{
"epoch": 7.253164556962025,
"grad_norm": 0.5584208200417776,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.1576,
"step": 69
},
{
"epoch": 7.3544303797468356,
"grad_norm": 0.4948162292093854,
"learning_rate": 4.180910442924312e-06,
"loss": 0.1002,
"step": 70
},
{
"epoch": 7.455696202531645,
"grad_norm": 0.6003769756361554,
"learning_rate": 4.033720411022235e-06,
"loss": 0.0634,
"step": 71
},
{
"epoch": 7.556962025316456,
"grad_norm": 0.632817072117396,
"learning_rate": 3.887395330218429e-06,
"loss": 0.1324,
"step": 72
},
{
"epoch": 7.658227848101266,
"grad_norm": 0.5512268097086501,
"learning_rate": 3.7420661812774577e-06,
"loss": 0.0783,
"step": 73
},
{
"epoch": 7.759493670886076,
"grad_norm": 0.5935696380747312,
"learning_rate": 3.5978630534699873e-06,
"loss": 0.1258,
"step": 74
},
{
"epoch": 7.860759493670886,
"grad_norm": 0.43674489606470374,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.1357,
"step": 75
},
{
"epoch": 7.962025316455696,
"grad_norm": 0.5040701068675334,
"learning_rate": 3.3133500630858507e-06,
"loss": 0.1221,
"step": 76
},
{
"epoch": 8.10126582278481,
"grad_norm": 0.5537439646466557,
"learning_rate": 3.173294878168025e-06,
"loss": 0.0895,
"step": 77
},
{
"epoch": 8.20253164556962,
"grad_norm": 0.3734160602926547,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.1119,
"step": 78
},
{
"epoch": 8.30379746835443,
"grad_norm": 0.36353809777994334,
"learning_rate": 2.8982138584521734e-06,
"loss": 0.0546,
"step": 79
},
{
"epoch": 8.405063291139241,
"grad_norm": 0.5432026494165365,
"learning_rate": 2.7634342584218364e-06,
"loss": 0.0701,
"step": 80
},
{
"epoch": 8.50632911392405,
"grad_norm": 0.7187854238009306,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.088,
"step": 81
},
{
"epoch": 8.60759493670886,
"grad_norm": 0.6028065371137555,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.0989,
"step": 82
},
{
"epoch": 8.708860759493671,
"grad_norm": 0.6029218748519589,
"learning_rate": 2.371581150947476e-06,
"loss": 0.1138,
"step": 83
},
{
"epoch": 8.810126582278482,
"grad_norm": 0.44030990822373384,
"learning_rate": 2.245515092739488e-06,
"loss": 0.0778,
"step": 84
},
{
"epoch": 8.91139240506329,
"grad_norm": 0.44206891285952626,
"learning_rate": 2.1219146715716332e-06,
"loss": 0.1197,
"step": 85
},
{
"epoch": 9.050632911392405,
"grad_norm": 0.4181481470988038,
"learning_rate": 2.0008905265604316e-06,
"loss": 0.0645,
"step": 86
},
{
"epoch": 9.151898734177216,
"grad_norm": 0.47866066389040807,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.0707,
"step": 87
},
{
"epoch": 9.253164556962025,
"grad_norm": 0.4855471053569369,
"learning_rate": 1.7670019939210025e-06,
"loss": 0.0592,
"step": 88
},
{
"epoch": 9.354430379746836,
"grad_norm": 0.42763865169197446,
"learning_rate": 1.6543469682057105e-06,
"loss": 0.107,
"step": 89
},
{
"epoch": 9.455696202531646,
"grad_norm": 0.3690131276831282,
"learning_rate": 1.544686755065677e-06,
"loss": 0.0515,
"step": 90
},
{
"epoch": 9.556962025316455,
"grad_norm": 0.3883137017573927,
"learning_rate": 1.438119515243277e-06,
"loss": 0.1044,
"step": 91
},
{
"epoch": 9.658227848101266,
"grad_norm": 0.7347299174387595,
"learning_rate": 1.3347406408508695e-06,
"loss": 0.0741,
"step": 92
},
{
"epoch": 9.759493670886076,
"grad_norm": 0.4812162548965024,
"learning_rate": 1.234642669981946e-06,
"loss": 0.1096,
"step": 93
},
{
"epoch": 9.860759493670885,
"grad_norm": 0.369367326407901,
"learning_rate": 1.137915203877003e-06,
"loss": 0.0738,
"step": 94
},
{
"epoch": 9.962025316455696,
"grad_norm": 0.48681431404942604,
"learning_rate": 1.044644826718295e-06,
"loss": 0.0798,
"step": 95
},
{
"epoch": 10.10126582278481,
"grad_norm": 0.38416364988319346,
"learning_rate": 9.549150281252633e-07,
"loss": 0.0772,
"step": 96
},
{
"epoch": 10.20253164556962,
"grad_norm": 0.2972730096642589,
"learning_rate": 8.688061284200266e-07,
"loss": 0.0706,
"step": 97
},
{
"epoch": 10.30379746835443,
"grad_norm": 0.3687593689489182,
"learning_rate": 7.863952067298042e-07,
"loss": 0.0854,
"step": 98
},
{
"epoch": 10.405063291139241,
"grad_norm": 0.28383029271802473,
"learning_rate": 7.077560319906696e-07,
"loss": 0.0463,
"step": 99
},
{
"epoch": 10.50632911392405,
"grad_norm": 0.329282008425657,
"learning_rate": 6.329589969143518e-07,
"loss": 0.0642,
"step": 100
},
{
"epoch": 10.60759493670886,
"grad_norm": 0.2500705047947425,
"learning_rate": 5.620710549772295e-07,
"loss": 0.0462,
"step": 101
},
{
"epoch": 10.708860759493671,
"grad_norm": 0.3989612983696247,
"learning_rate": 4.951556604879049e-07,
"loss": 0.1179,
"step": 102
},
{
"epoch": 10.810126582278482,
"grad_norm": 0.2662828923303931,
"learning_rate": 4.322727117869951e-07,
"loss": 0.0675,
"step": 103
},
{
"epoch": 10.91139240506329,
"grad_norm": 0.2711704696179231,
"learning_rate": 3.734784976300165e-07,
"loss": 0.0693,
"step": 104
},
{
"epoch": 11.050632911392405,
"grad_norm": 0.34511642958777133,
"learning_rate": 3.18825646801314e-07,
"loss": 0.0957,
"step": 105
},
{
"epoch": 11.151898734177216,
"grad_norm": 0.2889242607992141,
"learning_rate": 2.6836308100417874e-07,
"loss": 0.041,
"step": 106
},
{
"epoch": 11.253164556962025,
"grad_norm": 0.24450548412580397,
"learning_rate": 2.2213597106929608e-07,
"loss": 0.0564,
"step": 107
},
{
"epoch": 11.354430379746836,
"grad_norm": 0.35934542044635454,
"learning_rate": 1.801856965207338e-07,
"loss": 0.0804,
"step": 108
},
{
"epoch": 11.455696202531646,
"grad_norm": 0.26298606930327606,
"learning_rate": 1.4254980853566248e-07,
"loss": 0.041,
"step": 109
},
{
"epoch": 11.556962025316455,
"grad_norm": 0.31777226755411614,
"learning_rate": 1.0926199633097156e-07,
"loss": 0.0903,
"step": 110
},
{
"epoch": 11.658227848101266,
"grad_norm": 0.3193608086114228,
"learning_rate": 8.035205700685167e-08,
"loss": 0.0721,
"step": 111
},
{
"epoch": 11.759493670886076,
"grad_norm": 0.25486180463366104,
"learning_rate": 5.584586887435739e-08,
"loss": 0.0657,
"step": 112
},
{
"epoch": 11.860759493670885,
"grad_norm": 0.25091083569651457,
"learning_rate": 3.576536829081323e-08,
"loss": 0.078,
"step": 113
},
{
"epoch": 11.962025316455696,
"grad_norm": 0.25614891878084256,
"learning_rate": 2.012853002380466e-08,
"loss": 0.0652,
"step": 114
},
{
"epoch": 12.10126582278481,
"grad_norm": 0.3196686889199045,
"learning_rate": 8.949351161324227e-09,
"loss": 0.0842,
"step": 115
},
{
"epoch": 12.20253164556962,
"grad_norm": 0.306975262273726,
"learning_rate": 2.237838582483387e-09,
"loss": 0.0994,
"step": 116
},
{
"epoch": 12.30379746835443,
"grad_norm": 0.24537391909796794,
"learning_rate": 0.0,
"loss": 0.0314,
"step": 117
},
{
"epoch": 12.30379746835443,
"step": 117,
"total_flos": 4.902218749915955e+16,
"train_loss": 0.26980868625080484,
"train_runtime": 2812.7387,
"train_samples_per_second": 1.46,
"train_steps_per_second": 0.042
}
],
"logging_steps": 1,
"max_steps": 117,
"num_input_tokens_seen": 0,
"num_train_epochs": 13,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.902218749915955e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}